Preparation of titanium dioxide-double-walled carbon nanotubes and its application in flexible dye-sensitized solar cells

Abstract

Titanium dioxide-double-walled carbon nanotubes (TiO₂-DWCNTs) with DWCNTs/TiO₂ of 20 wt.% is prepared by a conventional sol-gel method. Doping the TiO₂-DWCNTs in TiO₂ photoanode, a flexible dye-sensitized solar cell (DSSC) is fabricated. The sample is characterized by scanning electron microscopy, X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR) absorption, ultraviolet-visible spectroscopy (UV-vis) absorption spectra,electrochemical impedance spectroscopy (EIS) technique and photovoltaic measurement. It is found that adding a certain amount of TiO₂-DWCNTs can efficiently decrease the resistance of charge transport, improve dye adsorption. Under an optimal condition, a flexible DSSC contained with 0.50 wt.% TiO₂-DWCNTs achieves a light-to-electric energy conversion efficiency of 3.89% under a simulate solar light irradiation of 100 mW·cm^-2.

Keywords

flexible dye-sensitized solar cell (DSSC) / titanium dioxide (TiO₂) / double-walled carbon nanotube (DWCNT) / sol-gel method

1 Introduction

Microbolometer is one kind of thermal infrared (IR) detector, which has a great development in recent years. It has a wide application, such as light emitting diode (LED) defect test, industrial product test and so on, due to its high performance and room temperature operation without cooling system. The incident infrared radiation is transformed into heat and temperature rise of the sensitive material causes physical property change of this material. The signal is detected by precision circuit. This is the principle of microbolometer.

Many materials can be used as sensitive materials for microbolometer, such as VO_x, which has a high negative temperature coefficient of resistance (TCR) in semiconductor materials, and Ti, which has a relative high positive TCR in metal materials [1,2]. Other materials like amorphous silicon [3], silicon diode [4], etc., also have been developed in recent years.

Porous silicon technology has been used in many fields such as particle filter [5], micro-hotplates [6] and thermal sensors [7]. The purpose of utilizing porous silicon for microbolometer is to form sacrificial layer to reduce thermal conductance. Comparing with other sacrificial layers, porous silicon has these advantages: easy to be removed quickly; compatible with deposition process of thin film and compatible with the process of standard IC.

In this work, we deposit VO_x thin film of high TCR about -3.5%/K. To test the property of this VO_x thin film, X-ray diffractometer (XRD) and atomic force microscopy (AFM) have been used. A research on the property of porous silicon has been investigated in detail. We integrate this film into microbolomer-based on porous silicon as sacrificial layer. The bolometer performance has been tested.

2 Vanadium oxide thin film of high TCR

As the common sensitive film, there are a lot of phases in vanadium oxide because of high TCR or transition from semiconductor to metal phase at a certain temperature. For example, vanadium oxide is mainly composed of V₂O₃, V₂O₅ and V₃O₄[2]. In this experiment, utilizing the method of pulsed DC magnetron sputtering, vanadium oxide thin film is deposited with vanadium metal target (99.99% purity) for 15 min at argon and oxygen mixed atmosphere under optimized fabrication condition. The movement of electrons is confined near certain cathode areas by magnetic field so that probability of the gas molecules collision with electrons increases. This improves the ionization events and then the deposition efficiency. To obtain a high quality of VO_x film, the ratio of argon gas and oxygen gas is controlled about 179 sccm∶20 sccm by precise mass flow controller with 0.78 A×397 V sputtering power. During deposition process, the substrate temperature of 250°C is relatively low, which is beneficial to on-chiped readout integrated circuit (ROIC) for large array microbolometer. The VO_x thin film sample for XRD test is deposited on silicon substrate with 300 nm Si₃N₄ buffer layer. Si₃N₄ film is deposited by plasma-enhanced chemical vapor deposition (PECVD) at 300°C. Semiconductor film resistance R(T) and TCR can be expressed as

(1)$$R(T)=R_{\mathrm{0}}\exp \left({\displaystyle \frac{\mathrm{\Delta }E}{K{T}^{\mathrm{2}}}}\right),$$

(2)$$\alpha ={\displaystyle \frac{\mathrm{d}R}{R\mathrm{d}T}}={\displaystyle \frac{\mathrm{\Delta }E}{K{T}^{\mathrm{2}}}},$$

where $\mathrm{\Delta }E$ is activation energy; K is Boltzmann constant; T is temperature; R₀ is resistance constant; α is TCR. The electrical properties of VO_x film is investigated by four-probe system with precise temperature controller. This VO_x thin film shows TCR of -3.5%/K and sheet resistance at room temperature is about 128 kΩ/sq shown in Fig. 1. It appears a good linear trend from 290 to 350 K without any phase transition phenomena. The typical XRD pattern of VO_x thin film is shown in Fig. 2. The spectrum shows peaks 1 (33.0°), 2 (52.5°), 3(55.3°) that are related to the reflection from planes V₂O₃(104), V₃O₅(510), VO₂ (211). No other significant peaks corresponding to other vanadium oxide are observed. The AFM photo of vanadium oxide is shown in Fig. 3. It indicates that thin film is well crystallized. The average height of crystallite is about 6 nm.

Fig.1 In(R) dependence on temperature for VOx thin film

Full size|PPT slide

Fig.2 XRD pattern of vanadium oxide

Full size|PPT slide

Fig.3 AFM micrographs of vanadium oxide

Full size|PPT slide

3 Preparation of porous silicon

Porous silicon is formed by the method [5-8] of anodic electrochemical etching of the silicon in the well-proportioned mixture of hydrofluoric acid (HF) and hydrogen peroxide (H₂O₂) solution. The charge exchanges between dopant silicon surface and electrolytic HF and H₂O₂ forms Schottky contact results in porous silicon [9]. Photoresist is used to pattern the needed region in order to make unmasked region pore. High dopant concentration p-type(111) silicon with a low resistivity of 2×10^-2 Ω·cm is used to increase current flux, so that silicon can be easily porous and dissolved. Bias current is applied by Pt electrodes in double reaction tank, while the silicon is fixed in the middle of insulated baffle.

Figure 4 shows that different concentration of HF solution has influence on corrosion rate of anodic oxidation while current is stable. With the same HF solution, the corrosion rate increases as the current increases. High concentration of HF solution has a faster corrosion rate than other low concentration of HF solution. With time passing by, the depth of porous silicon becomes deeper while keeping 30% concentration of HF solution stable as shown in Fig. 5. The depth of porous silicon shows linear dependence on anodizing time in 10 min.

The two channels of porous region are removed in KOH solution as shown in Fig. 6. KOH solution dissolves two channels unmasked region. Based on this quickly dissolved sacrificial layer (only two minutes), it is suitable for micromachining of the devices. A 2.5 μm cavity length of porous silicon (9 mA/cm² bias current, 2 min) corresponds to 8-14 μm atmosphere infrared transmittance spectrum because of λ/4 resonance theory [10].

Fig.4 Curves of corrosion rate of HF solution versus current density

Full size|PPT slide

Fig.5 Curves of depth of porous silicon versus anodizing time

Full size|PPT slide

Fig.6 Porous Si contour profile as sacrificial layer after dissolved

Full size|PPT slide

4 Process flow for microbolometer

Process flow steps are shown in Fig. 7. Firstly, the porous silicon sacrificial layer has been prepared by the method of anodic electrochemical etching as shown in Fig. 7(a). Si₃N₄/SiO₂/Si₃N₄(thickness: 200 /150/200 nm) “sandwich structure” supporting layer is deposited by PECVD (Fig. 7(b)). The sandwich structure plays roles of insulating, supporting, reducing heat-conductive and absorbing infrared. This kind of silicon dioxide/silicon nitride has advantage of well uniformity and hardness. Vanadium oxide sensitive film (150 nm) is deposited by utilizing the method of pulsed DC magnetron sputtering (Fig. 7(c)). The shape of this sensitive thin film is well patterned by wet etched method. On top of VO_x adjacent to the edge deposits a conductive layer by sputtering Cr/Au (60/100 nm). It sticks to the bottom layer well and shows a good conductivity. The conductive layer is formed by lift-off method (Fig. 7(d)). Then with the method of pulsed DC magnetron sputtering, controlling N₂, argon gas rates, TiN absorbing layer (300 nm) is deposited. PECVD Si₃N₄/SiO₂ (200/100 nm) makes further efforts to promote infrared absorption and to protect the bottom layers (Fig. 7(e)). Thus micro-bridge structure has been left to be formed. After the required shape and region have been etched to the bottom by MRIE (Fig. 7(f)), KOH solution immerses the sample only 2 min. H₂emerges from the solution during this anisotropic wet etching process. Finally, micro-bridge structure VO_x infrared bolometer unit has been investigated by scanning electron microscopy (SEM) as shown in Fig. 8. The minimum conductive layer width of micro-bridge is 5 μm. A wider metal layer (about 8 μm) connects the narrow one on one side, on the other side connects outer circuit. The VO_x sensitive region is of a circle shape whose diameter is 50 μm. All above is fully utilized for IR absorption and thermal conductance reduction.

Fig.7 Process flow of microbolometer based on porous silicon sacrificial layer. (a) Preparation of porous silicon layer; (b) deposition of supporting layer; (c) deposition and etching of VOx; (d) deposition of conduction layer; (e) PECVD films and deposition of TiN absorbing layer; (f) etching of absorbing structure; (g) removal of porous silicon sacrificial layer

Full size|PPT slide

Fig.8 Microphotography of microbolometer

Full size|PPT slide

Fig.9 Total absorption for microbolometer

Full size|PPT slide

5 Detector test

To analyze IR absorption for detector, Fourier transform infrared spectroscopy (FTIR) was used to measure the absorption. The curve is determined as the wavelength from 4-15 μm. The absorption of total films is shown in Fig. 9. The value of highest absorption at 9.87 μm reaches 98%. The average absorption of 8-14 μm is calculated to be about 74%. It means the total film system including TiN absorber is of good absorption for mid-infrared.

Fig.10 Test system for VOx bolometer

Full size|PPT slide

The vanadium oxide thin film mentioned above is compatible with CMOS technology. This thin film has been integrated into microbolometer. The bolometer sheet resistance is 128 kΩ/sq, with a TCR of -3.5%/K. The bolometer is packaged in a vacuum chamber. In front of chamber is an 8-14 μm optical window, which is biased by a DC power current source. The signal is measured through a low noise pre-amplifier and a lock in amplifier. Black body source of 850°C radiates the microbolometer through a chopper frequency which ranges from 5 to 800 Hz. The chamber is pumped and kept about 10 Pa. The bolometer test system is shown in Fig. 10. The purpose of utilizing porous silicon is to form sacrificial layer to reduce thermal conductance. The total thermal conductance is composed of thermal conductance of gas atmosphere G_gas, supporting legs G_leg and material radiation G_rad: G_total = G_leg+G_gas+G_rad. Because of such a low vacuum pressure, G_rad and G_gas can be ignored. So G_leg is the main thermal conductance for this bolometer. The thermal time constant $\tau$ with thermal conductance G of 4.8×10^-7 W/K was 5.4 ms, which is obtained through 3 dB frequency signal. The thermal mass C was about 2.59×10^-9 J/K with the relation: C = τ×G. The bolometer responsivity R_v and detectivity D* are given by

(3)$$R_{\mathrm{v}}={\displaystyle \frac{\alpha \eta I_{\mathrm{b}}{R}_{\mathrm{0}}}{g\sqrt{\mathrm{1}+{\omega }^{\mathrm{2}}{\tau }^{\mathrm{2}}}}},$$

(4)$$D^{*}={\displaystyle \frac{R_{\mathrm{v}}\sqrt{A_{\mathrm{d}}\mathrm{\Delta }f}}{V_{\mathrm{n}}}},$$

where α is TCR; η is infrared absorptivity; I_b is bias current applied to device; R₀ is bolometer resistance at room temperature; g is the thermal conductance; τ is thermal time constant; ω is modulation frequency. A_d is active absorptive area; V_n is noise; ∆f is frequency bandwidth. The curve of responsivity R_v at 9 Hz depended on chopper frequency is shown in Fig. 11. Detector’s responsivity has maximum value of 6.7×10⁴ V/W and it shows decline trend especially between 30-90 Hz. Comparing with our former bolometer based on porous silicon [11], thermal conductance has been reduced from the level of 10^-5 to 10^-7 W/K, due to the higher TCR of VO_x, the higher absorption and the less thermal conductance and less thermal mass. The responsivity has reached as 10 times as the former 5000 V/W. To reach the same responsivity, power consumption has only 1/10 than the former one because of the optimization of the structure, which means that the power consumption has been minimized greatly. The noise voltage remains a relative low level. Bolometer detectivity at 24 Hz with 9.8 μA bias current has maximum value of 1.09×10⁹ cm·Hz^1/2/W shown in Fig. 12.

Fig.11 Curve of responsivity (Rv) depended on frequency

Full size|PPT slide

Fig.12 Curves of detectivity (D*)and noise voltage versus frequency

Full size|PPT slide

6 Conclusions

Vanadium oxide thin film is deposited and VO_x film of TCR of -3.5%/K is obtained. Based on porous silicon as sacrificial layer, microbolometer of 74% IR absorptivity is fabricated to test the performance of VO_x thin film. The results indicate that this kind of bolometer integrated into this VO_x thin film, has maximum detectivity of 1.09×10⁹ cm·Hz^1/2/W at 24 Hz frequency and 9.8 μA bias current.

References

[1] O’Regan B, Grätzel M. A low-cost, high-efficiency solar cell based on dye-sensitized colloidal TiO₂ films. Nature, 1991, 353(24): 737–739

[2] Grätzel M. Recent advances in sensitized mesoscopic solar cells. Accounts of Chemical Research, 2009, 42(11): 1788–1798

[3] Lindstrom H, Holmberg A, Magnusson E, Malmqvist L, Hagfeldt A. A new method to make dye-sensitized nanocrystalline solar cells at room temperature. Journal of Photochemistry and Photobiology A: chemistry, 2001, 145(1–2): 107–112

[4] Longo C, Freitas J, DePaoli M A. Performance and stability of TiO₂/dye solar cells assembled with flexible electrodes and a polymer electrolyte. Journal of Photochemistry and Photobiology A: Chemistry, 2003, 159(1): 33–39

[5] Lindstrom H, Holmberg A, Magnusson E, Lindquist S E, Malmqvist L, Hagfeldt A. A new method for manufacturing nanostructured electrodes on plastic substrates. Nano Letters, 2001, 1(2): 97– 100

[6] Lee T Y, Alegaonkar P S, Yoo J B. Fabrication of dye sensitized solar cell using TiO₂ coated carbon nanotubes. Thin Solid Films, 2007, 515(12): 5131–5135

[7] Muduli S, Lee W, Dhas V, Mujawar S, Dubey M, Vijayamohanan K, Han S H, Ogale S. Enhanced conversion efficiency in dye-sensitized solar cells based on hydrothermally synthesized TiO₂/MWCNT nanocomposites. ACS Applied Materical and interfaces, 2009, 1(9): 2030–2035

[8] Ha B, Shin D H, Park J, Lee C J. Electronic structure and field emission properties of double-walled carbon nanotubes synthesized by hydrogen arc discharge. Journal of Physical Chemistry C, 2008, 112(2): 430–435

[9] Endo M, Muramatsu H, Hayashi T, Kim Y A, Terrones M, Dresselhaus M S. Nanotechnology: ‘Buckypaper’ from coaxial nanotubes. Nature, 2005, 433(7025): 476–478

[10] Cheng C X, Wu J H, Xiao Y M, Chen Y, Fan L Q, Huang M L, Lin J M, Wang J L, Tang Z Y, Yue G T. Polyvinyl pyrrolidone aided preparation of TiO₂ films used in flexible dye-sensitized solar cells. Electrochimica Acta, 2011, 56(21): 7256–7260

[11] Gao B, Chen G Z, Li G. Carbon nanotubes/titanium dioxide (CNTs/TiO₂) nanocomposites prepared by conventional and novel surfactant wrapping sol–gel methods exhibiting enhanced photocatalytic activity. Applied Catalysis B: Environmental, 2009, 89(3–4): 503–509

[12] Grätzel M. Perspectives for dye-sensitized nanocrystalline solar cells. Progress in Photovoltaics: Research and Applications, 2000, 8(1): 171–185

[13] Brozena A H, Moskowitz J, Shao B, Deng S, Liao H, Gaskell K J, Wang Y H. Outer wall selectively oxidized, water-soluble double-walled carbon nanotubes. Journal of the American Chemical Society, 2010, 132(11): 3932–3938

[14] Yu Y, Yu J C, Yu J G, Kwok Y C, Che Y K, Zhao J C, Ding L, Ge W K, Wong P K. Enhancement of photocatalytic activity of mesoporous TiO₂ by using carbon nanotubes Appl. Catal. A, 2005, 289(2): 186–196

[15] Wu J H, Li P J, Hao S C, Yang H X, Lan Z. A polyblend electrolyte(PVP/PEG+KI+I₂) for dye-sensitized nanocrystalline TiO₂ solar cells. Electrochimica Acta, 2007, 52(17): 5334–5338

[16] Jitianu A, Cacciaguerra T, Benoit R, Delpeux S, Béguin F, Bonnamy S. Synthesis and characterization of carbon nanotubes-TiO₂ nanocomposites. Carbon, 2004, 42(5–6): 1147–1151

[17] Yen C Y, Lin Y F, Hung C H, Tseng Y H, Ma C C, Chang M C, Shao H. The effects of synthesis procedures on the morphology and photocatalytic activity of multi-walled carbon nanotubes/TiO₂ nanocomposites. Nanotechnology, 2008, 19(4): 045604

[18] Yen C Y, Lin Y F, Liao S H, Weng C C, Huang C C, Hsiao Y H, Ma C C, Chang M C, Shao H, Tsai M C, Hsieh C K, Tsai C H, Weng F B. Preparation and properties of a carbon nanotube-based nanocomposite photoanode for dye-sensitized solar cells. Nanotechnology, 2008, 19(37): 375305

[19] Sawatsuk T, Chindaduang A, Sae-kung C, Pratontep S, Tumcharern G. Dye-sensitized solar cells based on TiO₂-MWCNTs composite electrodes: Performance improvement and their mechanisms. Diamond and Related Materials, 2009, 18(2–3): 524–527

[20] Lee K M, Hu C W, Chen H W, Ho K C. Incorporating carbon nanotube in a low-temperature fabrication process for dye-sensitized TiO₂ solar cells. Solar Energy Materials and Solar Cells, 2008, 92(12): 1628–1633

[21] Zhang D S, Downing J A, Knorr F J, McHale J L. Room-temperature preparation of nanocrystalline TiO₂ films and the influence of surface properties on dye-sensitized solar energy conversion. Journal of Physical Chemistry B, 2006, 110(43): 21890–21898

[22] Hsu C P, Lee K M, Huang J T W, Lin C Y, Lee C H, Wang L P, Tsai S Y, Ho K C. EIS analysis on low temperature fabrication of TiO₂ porous films for dye-sensitized solar cells. Electrochimica Acta, 2008, 53(25): 7514–7522

Acknowledgements

This work was supported by the National High Technology Research and Development Program of China (No. 2009AA03Z217), and the National Natural Science Foundation of China (Grant Nos. 90922028, 50842027).

RIGHTS & PERMISSIONS

2014 Higher Education Press and Springer-Verlag Berlin Heidelberg